Explaining Support for Authoritarianism in New Democracies

Transcription

1 University of Vermont UVM UVM College of Arts and Sciences College Honors Theses Undergraduate Theses 2015 Explaining Support for Authoritarianism in New Democracies Andrew G. Goss University of Vermont, andrewggoss@gmail.com Follow this and additional works at: Recommended Citation Goss, Andrew G., "Explaining Support for Authoritarianism in New Democracies" (2015). UVM College of Arts and Sciences College Honors Theses. Paper 18. This Undergraduate Thesis is brought to you for free and open access by the Undergraduate Theses at UVM. It has been accepted for inclusion in UVM College of Arts and Sciences College Honors Theses by an authorized administrator of UVM. For more information, please contact donna.omalley@uvm.edu.

2 Explaining Support for Authoritarianism in New Democracies Andrew Goss April 17,

3 Democracy is the worst form of government, except for all those other forms that have been tried from time to time. -Winston Churchill Chapter 1: Introduction In the last few decades, significant democratic advances have been made all across the world. A body of research on democratization has emerged to explain this phenomenon. However, one important question is often overlooked in the rush to try to explain the global democratic push: do the citizens of these democratizing countries actually want democracy? Often the answer seems to be yes. Perhaps because of universally accepted norms, positive connotations with the word itself, or dreams of economic success, democracy retains a large margin of support compared to authoritarian alternatives across the world, especially when investigated in simple terms. However, this support is not uniform and falters in some new democracies more than others. Surveys in several of these countries have revealed that surprisingly large constituencies still view an authoritarian government as preferable. This thesis will investigate the puzzle demonstrated in Chart 1.1, which shows the percentage of respondents across Latin America and Africa who agreed that authoritarianism, or a non-democratic regime, can be preferable to democracy. No country exceeded 30% support for authoritarianism good news for democracy but there is significant variance across different countries. Average authoritarian support in Latin America is 18.4% and ranges from 11.9% in Bolivia to 28.9% in Honduras. In Africa, authoritarian support is generally lower. The African average is 11%, less than even the lowest Latin American country. However, there is still notable variance across the region and Lesotho presents itself as a significant outlier with 29.7% of survey takers agreeing that an authoritarian regime could be preferable. The puzzle 2

4 that emerges in Chart 1.1 is: what explains the wide variation between countries in levels of popular support for authoritarianism? Chart 1.1: Authoritarian Support in Africa and Latin America Sources: Afrobarometer 2008/2009, Latinobarómetro 2011 The puzzle is also apparent when we examine patterns of authoritarian support in Central and Eastern Europe (CEE). To investigate this region, the best available source is the World Values Survey (WVS). We can gain a relevant glimpse into authoritarian arian support based on the percentage of survey takers who reject (view as fairly bad or very bad) both military rule and rule by a strong leader who does not have to deal with parliament or elections. In Romania and Ukraine, the level of rejection is extremely low: 18.8% and 27% respectively. In Poland and Estonia, the rate of rejection is higher: 56% and 58.6%. While the number of country cases is limited, it is convenient for analysis that the four countries divide neatly into categories of high and low authoritarian support. 3

5 Using similar measures of support for concrete regime alternatives in Latin America and Africa shows even greater variation in support for authoritarian regime types in these regions. 29.2% of respondents across 18 countries in Latin America agreed that they would support a military government in replacement of a democratic one if things got bad. Conditional support for military rule reached as high as 56.6% in Guatemala and as low as 9% in Costa Rica. In Africa, support for one-party rule reached 39.2%, 36.2%, and 34.1% in Lesotho, Mozambique and Burkina Faso respectively. The same measure found only 4.9% support in Senegal and less than a tenth in Madagascar. Appendix A lists these various measures of authoritarian support across the three regions observed in this study. These trends highlight the significant variation among different new democracies with regards to their support for the idea of an authoritarian regime. These measures of authoritarian support will serve as the dependent variable in the following investigation of the factors that might contribute to the differences observed. I will examine why authoritarian support remains a key feature in the political makeup of many of these countries despite the adoption of at least nominally democratic institutions. Beginning with the relevant literature, I will investigate factors that might be causing the persistence or emergence of support for non-democratic regime alternatives. Literature Review and Hypotheses The major thrust of academic work related to this particular issue has addressed the causes of support for democracy rather than support for authoritarianism. In many ways support for democracy (or democratic legitimacy) is the opposite side of the same coin and can offer clues for this project. However, there are important differences between these two measures of support. For example, a lack of democratic support could indicate widespread indifference to 4

6 regime type rather than an embrace of some concrete non-democratic alternative. Measures of support for authoritarianism also tend to show larger disparities between countries and can therefore guide a more careful and nuanced investigation. In this project, I am interested in investigating the root sources of popular support for the aim of reintroducing authoritarian institutions into newly democratic countries. Since the division between these two research questions is not well defined and often overlapping, the following review of the literature will draw on both measures of regime support in order to form hypotheses. Explanations for regime support tend to focus on economic, political and cultural factors. One classic line of scholarship has emphasized the role of economic factors in determining support for democracy. Writing in the 1950s Seymour Martin Lipset made the case that modernization would promote democratization. 1 His theory of modernization suggests that through industrialization, urbanization, and widening of education, democratic values would develop among the people. Later scholars have stressed the importance of the more immediate economic experience of new democracies in shaping attitudes toward democracy. 2 Proponents of this argument expect to find support for democracy in states that achieve growth and wealth under democratic institutions. Conversely, an experience of hardship and scarcity should decrease commitment to democracy. This school of thought helps generate the following hypothesis regarding support for authoritarianism: poor and/or stagnant economic performance under democratic institutions will cause disenchantment with democracy and a consequent preference for authoritarian rule. 1 Seymour Martin Lipset, Some Social Requisites of Democracy: Economic Development and Political Legitimacy, The American Political Science Review (1959), Adam Przeworski, Democracy and the Market: Political and Economic Reforms in Eastern Europe and Latin America (New York: Cambridge University Press, 1991); Herbert Kitschelt, The Formation of Party Systems in East Central Europe, Politics and Society (1992),

7 The investigation of these factors, carried out in Chapter 2, questions this hypothesis, drawing from and expanding on the literature presented above. I measure levels of education, urbanization, GDP per capita, GDP growth, inflation, economic inequality and poverty in countries with high popular authoritarian support as compared to those with lower support. Additionally, seeking to fully grasp the economic realities for citizens in these countries, I look at the effect of survey takers assessments of their countries and their own household economic situations on regime support. Many theorists warn against putting too much stock into a country s economic experience to explain support for democracy and authoritarian alternatives. Numerous studies looking at a wide variety of possible explanations suggest that political variables rather than economic variables are most important. 3 This body of research expects to find support for democracy and rejection of authoritarianism in states where democratic institutions have achieved rule of law, individual freedoms and rights, meaningful elections, and limits on corruption. Where this has failed, citizens will feel less committed to democracy and be more likely to embrace authoritarian options. The hypothesis that emerges from this literature is as follows: Failure to guarantee certain basic freedoms and securities associated with democratic institutions will drive citizens in new democracies to support some form of authoritarianism. Chapter 3 investigates this hypothesis and the importance of the political system for determining authoritarian support. Again comparing countries with high authoritarian support to those with lower authoritarian support, this chapter looks at general measurements of politics in 3 Geoffrey Evans, Stephen Whitefield, The Politics and Economics of Democratic Commitment: Support for Democracy in Transition Societies, British Journal of Political Science (1995), ; Richard Rose, William Mishler, Christian Haerpfer, Democracy and its Alternatives: Understanding Post-Communist Societies (Baltimore: John Hopkins University Press, 1998); Larry Diamond, Developing Democracy: Toward Consolidation, (Baltimore: The John Hopkins University Press, 1999). 6

8 each country as well as perceptions of corruption. Then, attempting to measure the experience of individuals in these countries, I look for correlations between measures of authoritarian support and several survey questions asking about democratic conditions on the ground. These include worsening/improvement of democracy, the delivery of various political goods associated with democracy, efficacy of the system, satisfaction with democracy, and corruption. Chapter 3 also investigates the effect of current and past political violence stemming from the regime in power. The hypothesis here is that an increase in repressive tactics used by the government under democratic institutions (particularly with respect to past levels of government violence) will lead some to be more open to a form of authoritarian rule similar to the seemingly safe past regime. Culture is another factor evoked to explain differences in regime commitment. Some scholars have suggested that cultural factors like Asian values or religious traditions like Islam can produce a population that is more subservient to authority and willing to live under an illiberal regime. At least one study, looking at South Korea, finds that certain aspects of Confucian cultural tradition drive support for authoritarian alternatives. 4 Conversely, using statistical analysis, Stephen Fish finds that devoutness of Muslims does not have a significant effect on support for democracy. 5 Proponents of the cultural argument would expect countries to be bound by their dominant culture. In general this means that support for democracy would be found in more liberal, Western countries while support for authoritarianism would be found in countries with more hierarchical or traditional cultures. We can simplify this discussion with the 4 Chong-Min Park and Doh Chull Shin, Do Asian Values Deter Popular Support For Democracy? The Case of South Korea, Asian Barometer Working Paper Series, No. 26 (2004). 5 Steven M. Fish, Are Muslims Distinctive? A Look at the Evidence (New York: Oxford University Press, 2011). 7

9 following hypothesis: countries that are not characterized by a liberal, Western culture will generally be more likely to embrace authoritarian alternatives. Using this hypothesis as a springboard, Chapter 4 attempts to quantify cultural differences between countries with varying levels of authoritarian support. I use Ronald Inglehart and Christian Welzel s index of emancipative values to evaluate the degree to which Western, pro-democratic culture has penetrated a given society. Additionally, I investigate the proportion of survey-takers who see goals of security and economic development as more of a priority for their country than deepening democracy and building responsive institutions. Finally, I compare levels of interpersonal trust across the new democracies observed. Overview and Methods The first section of this project will investigate the hypotheses enumerated above at the regional level for Africa, Latin America and Central and Eastern Europe. I will begin by examining the role of economics, politics, culture and history in the formation of preferences for authoritarian rule in these regions. The idea of focusing on the regional level has faced some recent resistance by those who believe that analyzing isolated regions could fail to capture a significant cross-regional pattern. 6 One could counter that a larger-scale analysis could similarly fail to identify a significant regional pattern. As a secondary inquiry, this project will seek to identify region-specific trends that may indicate that authoritarian support is better analyzed at this level. Country cases were selected (and labeled democracies) from among those included in the relevant surveys if they received a score of at least partly free on Freedom House s measure of Freedom in the World during the year of the survey used for analysis. They will be divided into 6 Yun-han Chu, Michael Bratton, Marta Lagos, Sandeep Shastri, Mark Tessler, Public Opinion and Democratic Legitimacy, Journal of Democracy (2008), 76. 8

10 two groups based on levels of authoritarian support. In Latin America, the first group contains countries with over 20% saying that sometimes an authoritarian regime is preferable. The second group is made up of the remaining states where less than 20% of those surveyed accepted authoritarianism under some circumstances. 7 Due to the abundance of democracy-related Afrobarometer questions and the comparatively small variation between cases, African countries were divided by a different measure of authoritarian support. Countries categorized as more willing to accept authoritarianism displayed less than 60% rejection of both strong-man and military rule as well as over 10% agreeing that sometimes an authoritarian regime can be better. All other African countries were considered less willing to accept authoritarianism. 8 For these two regions, I will seek to find significant differences between both groups of countries in terms of economic, political and cultural/historical variables. In Central and Eastern Europe, the four countries will simply be presented individually. During this regional analysis, I will also zoom in and scrutinize a specific country from each region with unusually high support for authoritarianism. These countries will be Guatemala, Lesotho and Ukraine. These cases were chosen because of both their persistent support for authoritarianism over several survey periods as well as their high support based on several different measures (See Appendix A). The second section of this project will investigate the peculiar case of high authoritarian support in Ukraine. I will then use the most similar systems (or method of difference) 7 Countries with high authoritarian support in Latin America: Brazil, Dominican Republic, Ecuador, Guatemala, Honduras, Panama, and Paraguay. Countries with low authoritarian support: Argentina, Bolivia, Chile, Colombia, Costa Rica, El Salvador, Mexico, Nicaragua, Peru, Uruguay, and Venezuela. 8 Countries with high authoritarian support in Africa: Burkina Faso, Lesotho, Malawi, Mali, Mozambique, Namibia, Nigeria, South Africa, and Tanzania. Countries with low authoritarian support: Benin, Botswana, Cape Verde, Ghana, Kenya, Liberia, Madagascar, Senegal, Uganda, and Zambia. 9

11 approach to root out the drivers of authoritarian sympathy in Guatemala as compared to Nicaragua. This process involves choosing countries that are similar in many aspects but have different levels of support for authoritarianism. Due to the likeness of the cases in this comparative approach, I should be able to identify important differences between the two countries that may influence regime support. The findings from the regional analysis will guide these smaller-scale investigations. Due to the fact that this investigation relies mainly on cross-national survey data, authoritarianism is conceptualized very generally as any non-democratic form of rule. Where survey questions simply ask respondents about authoritarianism as a regime choice, we should not expect all survey-takers to be applying the same definition. Often individual interpretations of what authoritarianism means will be determined by prior experience within the country or the experience of neighboring countries. Where possible, I will employ measures of support for concrete authoritarian alternatives where power is clearly not derived from the people, but rather from the military or a strongman ruler. Significance Many scholars have recognized the importance of legitimacy for the survival and development of any regime. Congruence theory is represented in a body of work arguing that political institutions in a country must be in line with similar orientations among the population. 9 That means that a democratic regime in a country with undemocratic or authoritarian citizens will be bound for failure. Larry Diamond makes this argument fairly convincingly in his book Developing Democracy: Towards Consolidation, where he asserts that for a democracy to 9 Christian Welzel and Ronald F. Inglehart, Political Culture, Mass Beliefs, and Value Change, Democratization, edited by Christian W. Haerpfer, Patrick Bernhagen, Ronald F. Inglehart, and Christian Welzel (New York: Oxford University Press, 2009),

12 consolidate, 70-75% of the population should consider democracy to be better than all other possible regime types. If this proves to be true, it would seem logical to look at the factors that drive this type of support. It would also seem particularly important to find countries that show wide support for those other possible regime types and understand why this is the case. However, there may be some reason to doubt the idea that popular preferences will be borne out in political institutions. For example, the trends observed in Chart 1.1 above show higher levels of support for authoritarianism in Latin America as compared to Africa even though levels of democracy are generally higher in the former region. Even if one does not buy the argument that authoritarian-minded citizens could be a danger for the development of democracy, this investigation still has value. The question remains disputed within the democratization literature. The persistence of support for authoritarianism is an interesting puzzle that can help improve an understanding of the way that countries transition to democracy and the role of the people in that process. This research project will attempt to add to what I believe to be an understudied but important factor in movements toward democracy. 11

13 Appendix A: Latinobarómetro (2011) Freedom House Score (2010) % Agreeing that under some circumstances, an authoritarian government can be preferable % Disagreeing that Democracy may have problems but it is still the best form of government % Agreeing I would support a military government in replacement of a democratic one if things got very bad Average Argentina 4 (F) Bolivia 6 (PF) Brazil 4 (F) Chile 2 (F) Colombia 7 (PF) Costa Rica 2 (F) Dominican 4 (F) Republic Ecuador 6 (PF) El Salvador 5 (F) Guatemala 8 (PF) Honduras 8 (PF) Mexico 6 (PF) Nicaragua 8 (PF) Panama 3 (F) Paraguay 6 (PF) Peru 5 (F) Uruguay 2 (F) Venezuela 10 (PF) World Values Survey ( ) % Saying Good: A Strong Leader who does not have to bother with parliament or Elections % Saying Good: Having the army rule % Saying democracy is not very important (1-4 on a 1-10 scale) Average Chile Colombia Ecuador Mexico Peru Uruguay

14 Afrobarometer (2008/2009) Freedom House Score (2008) % Approving of one-party rule % Approving of military rule % Approving of one-man rule % Agreeing that sometimes a nondemocratic government can be preferable Average Benin 4 (F) Botswana 4 (F) Burkina Faso 8 (PF) Cape Verde 2 (F) Ghana 3 (F) Kenya 6 (PF) Lesotho 5 (F) Liberia 8 (PF) Madagascar 6 (PF) Malawi 8 (PF) Mali 4 (F) Mozambique 7 (PF) Namibia 4 (F) Nigeria 8 (PF) Senegal 5 (F) South Africa 3 (F) Tanzania 7 (PF) Uganda 9 (PF) Zambia 8 (PF) Zimbabwe 13 (NF)

15 Afrobarometer (2010/2011) Freedom House Score (2011) % Approving of one-party rule % Approving of military rule % Approving of one-man rule % Agreeing that sometimes a nondemocratic government can be preferable Average Benin 4 (F) Botswana 5 (F) Burkina Faso 7 (PF) Burundi 10 (PF) Cape Verde 2 (F) Ghana 3 (F) Guinea 10 (PF) Kenya 7 (PF) Lesotho 6 (PF) Liberia 7 (PF) Madagascar 10 (PF) Malawi 7 (PF) Mali 5 (F) Mauritius 3 (F) Morocco 9 (PF) Mozambique 7 (PF) Namibia 4 (F) Niger 7 (PF) Nigeria 8 (PF) Senegal 6 (PF) Sierra Leone 6 (PF) South Africa 4 (F) Togo 9 (PF) Tunisia 7 (PF) Uganda 9 (PF) Tanzania 6 (PF) Zambia 7 (PF) World Values Survey % Approving of strongman rule % Approving of military rule % Disapproving of democracy Average Estonia Poland Romania Ukraine

16 Chapter 2: Economics and Regime Support An early line of scholarship, generally looking at the transitions to democracy in Central and Eastern Europe (CEE), stresses the importance of a new democracy s economic experience in shaping its population s regime preferences. Some proponents of this view highlight the importance of modernization-related factors like education and urbanization. Others tend to view regime preference as a simple economic consideration: would I (or my country) be better off economically under a different regime type? The expectation is to find support for democracy in states that achieve a net improvement in the economic situation of the country. In general, an experience of growth and wealth under democratic institutions should promote democratic support while an experience of hardship or scarcity should cause a desire to turn to some authoritarian alternative. Looking at the transitions in CEE, Adam Przeworski suggests that the populations of transitioning countries have little experience with a market economy and form expectations based on promises from political leaders. Their assessment of democracy as a regime then will be determined by how the reality of democracy and a market economy stack up to their expectations. 10 Herbert Kitschelt also offers a perspective on this issue in an article explaining the formation of party systems in CEE. He argues that countries that experience less economic development will have populations that are more authoritarian in their political orientations and preferred parties. Conversely, more economic development will produce a more libertarian, inclusive and participatory population Adam Przeworksi, Democracy and the Market (Cambridge: Cambridge University Press, 1991), Herbert Kitschelt. The Formation of Part Systems in East Central Europe, Politics and Society, 20 (1992),

17 Proponents of this line of thinking give a number of theoretical reasons for the connection between regime support and economics. These tend to take the form of modernization explanations and explanations rooted in more short-term economic experiences. Those viewing modernization as central to regime support claim that increasing affluence, a larger middle class, more free time, and better educational opportunities will democratize the population. As Kitschelt puts it, these conditions will increase citizens capacities to exhibit individualism and to claim greater control over their life sphere. 12 Others have argued that development and wealth provides citizens with a stake in the system, making politics less of a zero-sum game. 13 Those who find the recent economic experience of a country to be more important tend to fall closer to Przeworski s framework that regime support will be determined by how the reality of democracy stacks up to citizens expectations. Following this argument, citizens will judge democracy based on whether the government is able to provide an improvement in living standards or if, instead, economic conditions cause uncertainty and suffering. 14 Additionally, Economic performance in the short-term has been found to drive satisfaction with democracy, which is in turn correlated with support for democracy, suggesting at least an indirect effect on regime legitimacy. 15 Applying these ideas to authoritarian support, we may first expect to see high popular willingness to accept authoritarian institutions of government in less modernized countries characterized by lower per capita GDP, education and levels of urbanization. Perhaps in these 12 Ibid, Geoffrey Evans and Stephen Whitefield, The Politics and Economics of Democratic Commitment: Support for Democracy in Transition Societies, British Journal of Political Science, Vol. 25, No. 4 (1995), Ibid. 15 Larry Diamond, Developing Democracy: Toward Consolidation (Baltimore: John Hopkins University Press, 1999),

18 less developed countries, authoritarian values persist because they have not had the democratizing experience of industrialization, urbanization and education. Turning to short-term explanations, we might expect that in a situation of deprivation, high inflation, and poverty, citizens might blame the political system and shift their loyalty to some non-democratic alternative. 16 This assumes that citizens will view some kind of authoritarian regime as more stable, decisive and better armed to tackle difficult economic issues. Indeed the process of economic reform under democratic institutions sacrifices some degree of efficiency for a more inclusive, open decision-making process. In order to get a full picture of the differences in economic situations between countries, I look at both objective and subjective measures of how the economy is performing. First, I will investigate a series of objective (or macro-level) measures of both modernization and economic growth such as GDP per capita, GDP growth, urbanization, education, inflation, inequality and poverty, in order to capture each country s recent and historical economic experience under democracy. In the second half of this chapter I will look at several subjective (or micro-level) measures of perceptions of economic conditions derived from responses to survey questions contained in the Latinobarómetro, Afrobarometer and World Values Survey. Objective Measures The tables below include several measures that seek to capture important elements of the economic situation in the regions under investigation. To measure levels of modernization and development, I have employed levels of GDP per capita, percentage of the population that is urban, and mean years of education. All of these variables are expected to be lower in countries characterized by higher popular support for authoritarianism. To capture each country s more 16 Ibid. 17

19 recent economic experience under democratic institutions, I rely on ten-year averages of GDP growth and inflation. Lower GDP growth and higher inflation are expected to be associated with countries that display higher levels of authoritarian support, as citizens will be disappointed in the democratic government s inability to provide prosperity and may seek alternative means to do so. Related to this is a measure of the percent of the population below the national poverty line in each country. This seeks to capture the degree to which democracy appears able to tackle the difficult issue of its impoverished citizens. Higher poverty should be associated with a willingness to accept non-democratic solutions in the form of a more authoritarian regime. Poverty might also be linked to modernization and education levels, which could have an independent effect on regime support. The final measure employed below looks at levels of inequality captured by each state s GINI coefficient. 17 Although the scholarship does not generally indicate inequality as a driver of authoritarian or democratic support, there may be some reason to believe it is relevant. On the one hand, citizens might blame the current democratic government for failing to address this important economic issue, shifting their hopes to an authoritarian populist promising speedy reform. Alternatively, deepening democracy could threaten elite interests and turn some who are on the winning side of inequality toward support for authoritarian options. My analysis indicates a small but notable effect of inequality on authoritarian support in Latin America, but the drivers behind this trend are somewhat unclear. To determine whether these factors contribute to authoritarian support, the countries have been divided into two categories based on the criteria identified in Chapter 1. The following analysis will look for significant differences between the group of countries with populations 17 The GINI index measures how far the distribution of income in a particular country is from perfect equality. On a scale, a higher score indicates a more unequal country. 18

20 identified as more willing to support authoritarianism and those that are less willing. Additionally, I will separate out one interesting case in both Latin America and Africa that has notably high levels (Guatemala and Lesotho). The results suggest that there is significant regional variation with regard to the salience of economic factors. In Latin America, modernization and inequality appear somewhat relevant to explanations of authoritarian support. In CEE, poor short-term economic experience and low GDP per-capita characterize the countries with higher authoritarian support. In Africa, there is little evidence that economic factors are important at all. Latin America In Latin America, Table 2.1 shows some interesting differences between countries with higher authoritarian support and those with lower levels. In terms of modernization, the differences are as predicted with higher authoritarian support being present in countries with a smaller urban population, lower GDP per capita and lower levels of education. However, only urbanization is significant at the 0.1 level. In general, this offers some tentative support to the idea that countries that are slow to develop and industrialize may be prone to higher authoritarian support. An experience of urbanization could introduce more open, democratic ideas to a population as opposed to a more isolated and traditional rural environment. Still, a lack of development is not the death knell of popular authoritarian sentiment. Some countries, like Brazil and Panama, where authoritarian options are more popular, have levels of modernization comparable to there less authoritarian counterparts. Additionally, Bolivia and Nicaragua are characterized by relatively low authoritarian support despite some of the lowest levels of per capita GDP and urbanization. 19

21 Table 2.1 generally indicates little difference between the two sets of countries in terms of inflation, GDP growth and poverty level. This contradicts the predictions offered by the literature reviewed above. In this region, it appears that authoritarian support can thrive under both good and bad economic conditions. Based on this data, poor short-term economic performance does not seem to be fueling democratic defectors. Table 2.1: Economics and Authoritarian Support in Latin America Guatemala (With High Authoritarian Support) Countries With Higher Authoritarian Support (St. Dev.) N=7 Countries With Lower Authoritarian Support (St. Dev.) N=11 Modernization Urban Population 50% 64.14% 77.9% (% of Total) (12.44) (11.94) GDP Per Capita (USD) Mean Years of Schooling (Adults) Average GDP Growth Average Inflation* GINI Coefficient ~2011 Proportion Living Under National Poverty Line (3535) (1.40) (5327) 8.24 (1.32) Recent Economic Experience (1.14) (1.09) (2.83) (3.75) (0.15) 6.01 (2.64) (3.49).323 (0.11) Difference (two-tailed p- value) % (0.06) (0.18) (0.16) 0.70 (0.24) 1.49 (0.33) 4.95 (0.03).043 (0.54) *3 countries are missing data for this variable Sources: The World Bank, UN Human Development Reports, Latinobarómetro 2011 Moving beyond these basic economic indicators, there is an interesting difference between the GINI coefficients of both groups of countries indicating that, in Latin America, more unequal countries have populations that are more supportive of authoritarianism. This relationship is demonstrated in Chart 2.1 below. The two groups of countries have, on average, 20

22 about a 5-point difference on the GINI index. One possible interpretation of this is that the centrality of land reform in many emerging Latin American democracies combined with the difficulty of carrying it out under those new democratic institutions may cause populations to seek out alternative regime possibilities. 18 Alternatively, this gap could be caused by a fear among the entrenched elite that democracy might bring a more equal distribution of wealth and power. While I do not find decisive evidence in either direction, results from survey data presented further on seem to suggest gest that in some cases it is more likely the latter explanation that fits in Latin America. Chart 2.1: Inequality vs Willingness ness to Accept Authoritarianism in Latin America Sources: World Bank, Latinobarómetro Diamond, Developing Democracy,

23 Guatemala sits at the negative extreme of each measure of economic experience. It is characterized by an undeveloped economy, extremely high poverty and inequality, as well as relatively low GDP growth. Whether these extreme levels have a greater effect on regime support than the slighter differences observed across other countries will be explored with a closer look at survey data further on. In general I find that this is not the case. Africa In Africa, Table 2.2 shows no notable relationship between either modernization or economic growth indicators and authoritarian support. For all variables the difference between both groups was miniscule. In the cases of GDP per capita, GDP growth, inflation and poverty the data actually shows poorer conditions among the group of countries less supportive of authoritarianism. Lesotho is somewhat unique in that it has lower GDP growth and greater inequality than the regional average. Again, the degree to which this could indicate that relatively extreme economic stagnation has a greater effect on regime preference than smaller differences between other countries will be observed using subjective measures. However, the data as a whole seems to contradict the idea that economic factors play a role in forming opinions about authoritarian regime alternatives in Africa. This is mostly consistent with the findings of similar studies on the region Michael Bratton and Robert Mattes, Support for Democracy in Africa: Intrinsic or Instrumental? British Journal of Political Science 31(3) (2001)

24 Table 2.2: Economics and Authoritarian Support in Africa Variable Lesotho (With High Support for Authoritarianism) Countries With Populations more Willing to Accept Authoritarianism (St. Dev.) N=9 Urban Population (% of Total) GDP Per Capita (USD) Mean Years of Schooling (Adults) Average GDP Growth Average Inflation * GINI coefficient ~2006*# Proportion Living Under National Poverty Line 25% 35% (13.74) (2416) (2.51) Countries With Populations less Willing to Accept Authoritarianism (St. Dev.) N=11 Modernization 41.3% (14.49) 1937 (2136) 5.43 (1.75) Recent Economic Experience (1.64) (4.30) (4.22) (12.05) (0.13) *excluding zimbabwe #data used was the closest to 2006 available Sources: Afrobarometer 2008/2009, The World Bank 9.18 (5.85) (7.51).461 (0.18) Difference (p-value) -6.3% (0.36) (0.84) (0.52) 1.12 (0.45) (0.58) 1.81 (0.71) (0.86) Central and Eastern Europe While it is difficult to make any definite conclusions based on the small number of cases available for Central and Eastern Europe, the table below reveals some interesting trends. The more authoritarian-leaning cases (Ukraine and Romania) demonstrate notably worse economic performance than their less-authoritarian counterparts (Estonia and Poland). Ukraine and Romania have lower average GDP growth over the ten years prior to the survey, a higher ten- 23

25 year average of inflation and a higher proportion of the population living under the national poverty line. These findings are in line with the expectations of previous scholars looking at the region and suggest that an experience of economic hardship might lead some to reject democracy for some illiberal alternative. Still, a 0.3 difference (at the highest) between GDP growths in these countries does not suggest a huge effect of economics on regime preference. Oddly, Inequality was actually higher in the cases with less support for authoritarianism. Table 2.3: Economics and Authoritarian Support in Central and Eastern Europe More Willing to Accept Authoritarian Alternatives Less Willing to Accept Authoritarian Alternatives Country (% Rejection of both one-man and military rule) Ukraine (27% rejection) Romania (18.8% rejection) Poland (56% rejection) Estonia (58.6% rejection) Modernization Urban Population 69% 54% 61% 68% (% of Total) GDP Per Capita ,648 18,783 (USD) Mean Years of Schooling (Adults) Average GDP Growth ( ) Average Inflation ( ) GINI Index (2010) Proportion Living Under National Poverty Line Recent Economic Experience Sources: World Values Survey , The World Bank Levels of education and urbanization are generally similar across all four countries. However, Ukraine and Romania have significantly lower GDP per capita than Estonia and Poland. This does not offer strong evidence that modernization is responsible for authoritarian 24

26 support in this region but we certainly can t rule out the idea that relative levels of wealth have some effect. Ukraine is more distinct than Romania for almost all of the variables except for GDP growth, which is only slightly larger than growth in Romania. Although this data does not allow any definitive conclusions, there is certainly some preliminary evidence that economic difficulty may drive an embrace of authoritarianism in Central and Eastern Europe. The degree to which these differences are mediated through citizens perceptions of the economy will be explored in the next section. Subjective Measures The variables used above capture the economic situation of the countries as a whole but sometimes these numbers can be deceiving or ignore the real day-to-day experience of individuals in the country. For this reason, many authors investigating the issue of regime support have opted to use survey data to determine each individual s personal experience with the economy rather than using broad macro-level data. 20 This approach assumes that changing regime preferences will be mediated through an individual s perception of the economic situation rather than the official numbers. The following tables investigate several subjective, personal measures of economic growth. For each region I will list Kendall s tau-b correlations between authoritarian support and economic factors. The survey questions used for each region can be found in Appendix B. I have also included the country-specific data for Lesotho and Guatemala in addition to data from the region as a whole. This will help explore the possibility, suggested above, that extremely poor economic conditions could cause a bump in authoritarian support even though 20 Evans and Whitefield; Richard Rose et al., Democracy and it s Alternatives, The John Hopkins University Press: Baltimore (1998). 25

27 middling conditions do not have the same effect. The findings are generally consistent with those for the objective variables, showing a limited effect of economic variables that differs from region to region. The analysis of exceptional cases appears to reject the idea that notably poor conditions experienced in Guatemala, Lesotho and Ukraine have a greater effect on authoritarian support. Latin America Attempting to capture similar variables to those used above, I have included both shortterm and long-term measures of perceptions of the economic situation in Latin America. I look at perceptions of survey-takers own as well as their countries economic situations. Also investigated are perceptions of fairness of income distribution, the government s ability to handle poverty, and education levels. The correlations of these variables with various measures of regime support are presented below in Table 2.4. The results show that, while many elements of an individual s perception of their own and their country s economic situation are related to regime support (and generally in the predicted directions), the effects are small. Citizens who view the country s economic situation poorly are more likely to prefer authoritarianism, accept military rule and believe that democracy is not always the best form of rule. Among generally weak predictors of authoritarian support, the most significant appear to be levels of education and perceptions that the government cannot solve the problem of poverty. Again, these relationships are notably small but they are most powerful for explaining differences in the Churchillian notion that democracy may have problems but it is the best system of government. Economically frustrated Latin Americans seem comparatively reluctant to endorse more concrete forms of authoritarian rule. 26

28 Interestingly, perceptions of unfair distribution of income are not connected with all measures of support for authoritarianism. This may help rule out the idea that the poor, dissatisfied with an unequal system, are clamoring for authoritarian rule. On the contrary, the case of Guatemala may offer some evidence that it is actually the wealthy that tend to find authoritarianism suitable. Lower levels of education are significantly correlated with a decreasing preference for authoritarianism in Guatemala, the reverse of what the modernization literature might predict. However, if we consider the fact that the poor are much more likely to lack education in Guatemala, we may find an explanation to this apparent contradiction in another body of work. Carles Boix has argued that for the democratic game to work in a given country, it is crucial that there be some degree of equality of conditions. Absent this level playing ground, the incentives to cheat become irresistible. 21 This means that those who benefit most from inequality might be open to authoritarian forms of rule that could prolong their position of strength. It seems entirely possible that this is operating in Guatemala (a highly unequal country), but the degree of inequality s overarching relevance for authoritarian support in the region as a whole remains unclear. Additionally, I find evidence in Chapter 6 that the division in authoritarian support in Guatemala is more clearly tied to ethnicity than class. 21 Carles Boix, The Roots of Democracy, Policy Review (February & March 2006). 27

29 Table 2.4: Subjective Economic Assessments and Regime Support in Latin America Country s Economic Situation Bad Personal/Family Economic Situation Bad Negative Change in Personal/Family Economic Situation Income Distribution Unfair Government Can t Solve the Problem of Poverty Lower level of Education Attained * significant at.05 level ** significant at.01 level Authoritarianism can be preferable Support military rule Disagree that democracy is better than alternatives.035**.036**.136** ** ** **.014*.086** **.062** **.046**.140** **.067**.066** -.110** Authoritarianism can be preferable (Guatemala) Perhaps the most surprising finding here is the fact that no subjective measure of the economic situation helps explain authoritarian support in Guatemala. While it is important to point out that the much larger number of cases for the entire region would be affecting statistical significance, the correlations are quite low as well in Guatemala. Only education (an aspect of modernization) seems to be related to levels of regime support among the population. This seems to run contrary to the idea, suggested above, that the extremely poor economic conditions in Guatemala might make the democratic government s economic performance all the more important for garnering legitimacy. Instead, we see that this kind of performance is irrelevant to regime support in the country. 28

30 Africa In Africa, a slightly different set of economic variables was used based on the available survey data. Similar to Latin America, this analysis includes perceptions of both the country s and one s personal economic situation. Also used is a measure of how the survey-taker feels their economic situation stacks up to that of other citizens. To capture poverty, I have included a question asking how often the respondent s family goes without food in a 12-month period. This is different from the question used in Latin America that captures feelings about the government s capability to alleviate poverty rather than the respondent s own situation. Finally, to gauge the effect of modernization I have employed levels of education and a simple rural/urban variable for where the survey took place. Again, the specific measures used below can be found in Appendix B. Consistent with the results from objective economic measures, there seems to be less of an impact of economic assessment on regime support in Africa compared to Latin America. Most economic measures are statistically significant for explaining some (but not all) measures of authoritarian support. However, their effect is very small. Only poverty and education levels consistently explain regime support and these too have limited impacts. In Lesotho, short-term economic evaluations as well as impoverished conditions do not help explain the high levels of support for authoritarianism in the country. Modernization indicators, however, do seem to impact regime legitimacy. Levels of education and the rural/urban measure are both statistically significant with correlation coefficients comparable to the highest ones found in the region as a whole. 29

31 Table 2.5: Subjective Economic Assessments and Regime Support in Africa Country s Economic Situation Bad Personal Economic Situation Bad Personal Economic Situation Better than other citizens Family Often Doesn t Have Enough Food Level of Rejection of Authoritarianism Approval of One-Party Rule Approval of Military Rule **.046** **.049** ** ** **.037**.020** * -.114** -.080**.126** Education Rural Area **.025** -.085** * significant at.05 level ** significant at.01 level Rejection of Authoritarianism (Lesotho) The data here suggests that we accept some limited impact of economic factors on authoritarian support in Africa. However, this effect is very small, especially compared with the other regions under observation. Only education and poverty help explain all three measures of authoritarian support included in this analysis, albeit with limited predictive power. Central and Eastern Europe To measure the effect of the economy on authoritarian support in Central and Eastern Europe I employed questions included in the World Values Survey. To obtain an idea of overall contentment with living conditions, a measure of satisfaction with one s personal financial situation is included below. Poverty is measured by the respondent s answer to how often they 30